◦ Developing a wireless readout system for passive interrogation of soil sensor, using the USRP B200 mini to transmit a small interrogating signal that interacts with the sensor’s resonator and receives a response for soil condition analysis.◦ Aiming to build a stand-alone solution that autonomously interprets soil conditions based on sensor data and deploythis system on drones for efficient monitoring of soil health over large agricultural areas to support precision farming.

◦ Worked under the supervision of Dr. Zheng Chen on the project titled, “Decentralized Stochastic Optimization using Gossips and Compress Communication”, focusing on stochastic optimization in a multi-worker setting.◦ Proposed a broadcast method based on random activation of the nodes in a decentralized stochastic optimizationsetting to imitate a real life scenario where there may not be perfect communication between the nodes.◦ Explored and implemented various quantization techniques to enhance compressed communication, helping tooptimize the trade-off between accuracy and communication costs in distributed learning environments.◦ Skills: Compressed Communication, Gossips Algorithm, Decentralized Stochastic Gradient Descent

◦ Worked under the supervision of Dr. Ping Wang to develop an efficient caching policy using Proximal Policy Optimization (PPO) in an IoT Network based scenario for transient data, which helps meet the standardquality of service requirement while bypassing the limitations of the IoT network.◦ Formalized and implemented the smart caching algorithm in software (python) and compared with pre-existing works, increasing the caching efficiency of the edge node by over 50% compared to currently used LRU and LFUmethods.◦ Extension of this work was presented at the International Conference on Communications in China (ICCC), 2023 held in Dailan, China and another has been submitted to the IEEE Transactions on Vehicular Technology.

◦ Worked on the project titled, “Joint parameter estimation and symbol detection in OFDM-based IoTNetwork: Machine learning-based approaches”, under the guidance of Dr. Amrita Mishra. This project wassponsored by the Science and Engineering Research Boards (SERB), Department of Science & Technology,Government of India.◦ Applied the Expectation-Maximization algorithm for semi-blind wireless channel estimation in ambient backscatter systems.◦ Aimed to enhance prior work by transitioning from blind to semi-blind channel estimation techniques, improving accuracy and performance.◦ Skills: Expectation-Maximization, Wireless Communication, Coding-Decoding, Channel Estimation